A Probabilistic Network-based Model Approach for the Development of Efficient Epidemic Strategies for the City of Chanute, Kansas
People
PIs: Pietro Poggi-Corradini (Mathematics), Walter Schumm (Family Studies), Caterina Scoglio (ECE), Steve Kubler (City of Chanute)
Student: Phillip Schumm (ECE).
Abstract
This research task will provide policy makers with information that will help them create better policies for mitigating outbreaks of infectious diseases. We will study the conditions for the infection spreading to be predictable and largely based on the topology of the contact network. We will study where and how to target subpopulations or communities by identifying either the best geographical location for public relief stations or the most efficient distribution of vaccine vouchers. Additionally we will explore social contact dynamics as a behavioral mitigation of potential epidemics.
A survey of the general public was undertaken in Neosho County, using a combination of survey techniques, following the tailored design method in terms of personalization and use of multiple follow-ups but without the use of financial incentives, stamped return envelopes, graphics or photographs, or a prenotice letter. An overall response rate of 65% was obtained, though a response rate of nearly 75% was obtained from residents of an urban location (population 9,119; occupied households 3,720 in 2010) compared to only 56% from two small rural towns (populations 497/126; occupied households 197/58 in 2010). Response rates compared favorably to those obtained by the U.S. Census at those locations and to response rates obtained using a larger array of techniques from the tailored design method. Even when research budget limitations require modifications of the tailored design method, high responses rates from the general public, at least in selected rural areas, may still be obtained.
We have developed an exact simulation of the classical SEIR model which classifies each individual as Susceptible (S), Exposed or Latent (E), Infectious (I), or Recovered (R). The movement transitions from S to E to I to R are tracked as events arriving according to a Poisson processes with the different rates of transition. With this simulator, we have begun to quantify the epidemic situation considering no mitigation, population-based vaccine distribution, individual-based vaccine distribution, and location-based vaccine distribution. These processes are simulated upon a directed, asymmetric, and weighted contact structure designed from the survey responses.
Presentations & Publications
Walter R. Schumm, Caterina Scoglio, & Vance Theodore, "ARE HIGH RESPONSE RATES POSSIBLE FROM SURVEYS OF THE RURAL GENERAL PUBLIC WHEN RESEARCH BUDGETS ARE LIMITED?" In preparation. 2011
